[Previous] Example: Effective Altruism Is Incompetent | Home | [Next] Politics Discussion

Three Discussions Approach

This post explains a way of organizing a discussion. It’s meant to be useful in some cases, not all the time. It doesn’t require that the other person know it’s being used. This method can be collaborative, or can be used as tips to guide your own actions.

The problem: people debate endlessly without anything being resolved.

The method: instead of having many debates, focus on reaching clear conclusions about three issues.

How? Ask the other person what they think is important or interesting. Get them to say something serious about an important issue (or link something they already wrote). Then focus on that. Instead of discussing whatever they carelessly say mid-discussion, try to get something more substantial that you can reply to. (People shouldn’t say careless things in discussions that they wouldn’t take responsibility for … that’s irresponsible … but they do.). Then clearly point out mistakes. Do one issue at a time, three times.

Tips: Focus criticism on key topics, not tangents or cherrypicked errors. Preferably, the criticisms to will point out important problems, not “this is incomplete” or “this is sloppy” (that doesn’t mean ignoring incompleteness or sloppiness, it means trying to get them to provide material which is more complete and effortful so that you have something good to respond to). If they can’t produce anything good (in your opinion), get them to say they think something they wrote is good (in their opinion), then point out that it’s incomplete or sloppy. That means they’re a poor judge of quality (which is an important criticism, but that should be your backup plan only if you can’t get them to say anything decent about a primary topic of interest like dinosaurs, history, politics, physics, etc.).

People make lots of excuses about their errors. They don’t want to figure out what caused their error (often a bad thinking method or static meme) and what other errors that cause could cause (often lots) and then fix the underlying problem. Focusing on three high quality critical interactions can reduce excuses. Pick things where they’d have few excuses for being wrong.

Does this method assume you’ll be doing all the criticizing, and unfairly have them stick their neck out while you don’t? That depends. People are welcome to criticize any of my important pieces of writing. Their criticism can be two of the three issues discussed, but shouldn’t be all three. If you want to use this method but have no public writing available for anyone to criticize, that’s a problem.

The second problem: People want answers to their questions, and corrections of their errors, one by one. What they should be doing is learning how to think better for themselves – learn better thinking methods, critical thinking skills and philosophy – so that they can answer more of their own questions and correct more of their own errors. Often, people want to ask a bunch of questions while not saying anything substantial themselves, so they minimize the ideas they expose to criticism.

It’s inefficient to outsource your thinking to me or to another wise person. I don’t have time to answer all your questions. I’ll answer a few if I like them, and to see if you learn much, and because people are interesting, and to interest people in learning my thinking methods (by giving examples of my wisdom). But most of your questions have to be answered by creating your own thinking and research skills, not by using mine. It’s the difference between teaching a man to fish and doing the fishing for him then giving him fish. Answering a question is giving someone a fish.

The solution: People should take an interest in learning to be better thinkers so they make fewer errors and can effectively find or create good ideas. I’ve written (and talked) a lot about how to do this and I’m open to questions about it.

Part of what learning involves is changing one’s mindset. It’s one thing to be a peer or equal, who is contributing about as many fish as he receives. It’s another to be unable to fish. (Or maybe you can only catch small fish, but you ask questions and make claims about big fish and are talking to people who know how to catch big fish. Big fish are complicated ideas.) You need to know which situation you’re in and act accordingly. Should you focus on learning more (to catch up to existing knowledge), or should you pursue your directly projects (with critical discussions and learning being secondary)?

The third problem: People view themselves as peers when they should be learners. And they don’t want to change that. They think they already are educated, good thinkers who can catch their own fish. They view their questions (areas of ignorance) and errors as occasional things, not a major pattern.

The solution: Show them their errors using three clear examples. Show them their ability to deal with ideas is less effective than they think it is, or less effective than it could be if they had the thinking skill that you do. Show them that you catch substantially bigger fish than they can, which is a skill they should learn if they want to successfully contribute anything important to human knowledge (or want to fight with their family less, or otherwise have a better life).

The three discussions method serves multiple purposes. It helps clarify the outcomes of discussions, and it helps limit how many different discussions happen before the patterns in the discussions are addressed, and it helps clarify the relative skill and knowledge of the participants, and it helps show people why they should try to learn to think better (because, three out of three times, they missed errors in their ideas – the type of errors that make the difference between success and failure).

The three discussions method can also show when something else is going on. Maybe the other guy will be right on some of the issues. Maybe there won’t be a pattern of error. Maybe he is wise, too. Maybe he can contribute a fair amount. That’s possible. Maybe you’re roughly equals. Maybe he knows far more than you, and you should be trying to learn from him. The three discussions method helps find out what the situation in a time-and-effort-efficient way.


Link: my discussion forums.

The first comment, below, is a second article on this topic with more info.


Elliot Temple on May 1, 2019

Messages (1)

Different Article Version

This post explains a way of having a discussion which I think will be good to use with many new people interested in FI. This is one way of organizing a discussion which I think will be useful; it’s not *the* way. It’s a tool to be used sometimes.

What issue does it address? Many people think they know more than they do. They don’t understand the gap from their knowledge to the sort of high quality knowledge FI consists of. They only learn FI stuff superficially, which is all they realize is possible, and never get very far. They need to start building up their critical thinking skills in a progression, starting with much more basic stuff where their error rate is lower, and practicing each skill until they can do it with low attention and a high success rate, before building much on it.

People disagree with me about their level of skill and knowledge compared to mine. They want demonstrations (examples of them being wrong, confused, low skill, ignorant of background knowledge, etc.). After they see that, they will maybe consider trying to learn the basics better as I advise. But I find that, after a dozen demonstrations where they were mistaken, they want more demonstrations. It’d be better to create three clear and high quality demonstrations instead of a dozen or more less clear demonstrations. Meanwhile, I think I should stop engaging in other discussions.

What does this mean? I ask them to *say three things which are their best work*. No excuses about being tired, being distracted, or it not being the topic they most care about. Three things where they’d be surprised to be mistaken, and would really appreciate to discover they were making any mistakes. The three things should be done *one at a time*. If all three times they are decisively refuted, and clearly were mistaken, then that should be enough examples and they should start focusing on learning. They should, at that point, accept that their skill is too far below me to debate me, without me having to tell them why they’re wrong, in detail, in every new case.

(Actually, it’d be even better if it was just pretty good work, not their best. If you have to try your very best to keep up with a discussion, it’s too hard for you. You need to up your skills so you can keep up without it taking so much work. Otherwise you’ll rarely participate because it’s too tiring. And the need to try hard means you haven’t mastered what you’re doing. If it’s so hard, you should be practicing to get better at it, not thinking you’re done learning.)

The three high-quality, high-effort things, which they expose to criticism, should be in writing. Video or audio is OK (but not preferred) if you provide a written transcript to enable replying to exact quotes.

What should the things be like? Original ideas are fine. Explanations of your understanding of an important issue, even if totally unoriginal, are fine. Criticism of something important that I wrote is fine too. If they want to take on the role of critic, instead of exposing positive claims to criticism, they can do that.

In each thing, you should say something substantial. I don’t want to be criticizing “This doesn’t really say much”. That’s not ideal. But if you really think you said something substantial, and it’s all fluff, that will serve as a demonstration of your confusion and error.

Each of the three things will be discussed to a conclusion or impasse. If they go silent, I shouldn’t talk to them about other things. They can either continue the discussion or explain why not. It’s important to state what the impasse for a discussion is. That let’s people understand why the discussion is stopping, and judge or criticize that reasoning. If you stop discussing one of your three things without stating any reason, people should reasonably assume you’re being irrational. If you will explain a reason, then it allows for other judgments.

If the impasse is “Joe won’t talk anymore because he hates being wrong and it started to look like his initial position might be refuted”, it’s good to know that’s what happened. If that’s not what happened, Joe can explain. If it did happen and Joe wants to lie, he may get caught in a lie. Lying is a lot harder to do right than silence. Our culture currently doesn’t expect people to even try to explain *why* they are ending a discussion, even one that involved a lot of time and effort by other parties. That ought to be expected. It’ll make irrationality and bias harder to get away with. It’s hard to come up with a *good*, rational reason to end a discussion if your actual reason is that you stopped talking because you couldn’t come up with refutations of some of the other guy’s arguments.

Next step: I will attempt to point out your mistakes for each thing. If you can keep up in the discussion and it’s reasonably equal, great, much respect. It will have been good that we found that out and we can have many more discussions going forward. If you win one debate, I will like you, read your blog and ask you questions.

If I can clearly and decisively show you were confused in 3 out of 3 discussions, then you will concede a large skill and knowledge gap, and stop trying to act like my peer or equal, and will start trying to learn in the ways I’ve explained elsewhere. In short, you need to start with stuff where your error rate is very low and build on that. For each new thing, you need to get it correct first with high attention, effort and detail, then practice it until it’s easier. Once you can do it with a very low error rate without much effort, then you’re ready to build on it. You can’t build up a complex knowledge structure with high-error building blocks. Learning requires not merely reading things, but talking about them and exposing your understanding to external criticism. And if you’re making lots of mistakes (which you should listen to my judgment on if you were decisively wrong in 3 out of 3 serious discussions), you need to read easier stuff and practice more basic skills and work your way up. Working on stuff where your error rate is high, and many of the errors are because your error rate as prerequisite skills is also high, is not an effective way to make progress.

What if the discussions have unclear results? We will both work to clarify them instead of just giving up and ignoring the issue. If we *both* agree, we can drop one of the discussions (agree it had an unclear result) and replace it with a new thing. I will try to organize the discussion for clarity of outcome. That’ll be primarily my job.

What if I point out “small” errors? The test of this is: can you easily fix them? If I point out a typo, you can say “OK, I retract that sentence and replace it with this other one instead…” You can revise what you said to address minor issues. But if you’d have to make rewrites to what you said in order to fix it, then you were wrong in a significant way.

In summary, I think it’s important to create some clarity about whether a non-humble new person is on my level or not. Unacknowledged or denied large skill gaps cause conversational problems (like a 5th grader trying to debate a physicist, or make suggestions, without realizing that he hasn't read any books on physics, and doesn't know algebra, and those facts matter). Should they share their thoughts and ask complicated questions like they know what they’re talking about and can contribute anything, or should they focus on learning more so that they could become good enough to contribute in the future? We can test that and create knowledge about it by having three discussions, where they do their best (no excuses). If I can clearly, decisively point out significant errors and confusions in all three cases, then we have our answer. And if I can’t, that is also important to find out.


curi at 3:38 PM on April 30, 2019 | #12250 | reply | quote

Want to discuss this? Join my forum.

(Due to multi-year, sustained harassment from David Deutsch and his fans, commenting here requires an account. Accounts are not publicly available. Discussion info.)